Return migration is of special significance in the study of migrant selectivity. It is plausible to assume that much, but by no means all, of return migration serves to "weed out" the less successful migrants. As a proposition: Return migrants from large cities on the whole will be negatively selective from the total group of migrants, thus increasing the positive selectivity of those who remain in the cities. The difficulties of testing this proposition are formidable, but it is crucial for migrant-nonmigrant comparisons.
SPECIFIC CHARACTERISTICS OF MIGRANT SELECTIVITY
Let us consider the specific ways in which migrants are selected from the populations from which they originate. A number of comparisons are possible, but the choices will be made in part to indicate the variety of relevant variables. We begin with three conventional demographic variables (age, sex, and marital status), proceed with two socioeconomic ones (education and occupation), and end with a psychological variable (risk-taking propensity).
More than 30 years ago, Thomas (30) in her review of the existing literature on migration differentials—interestingly enough, it was limited almost entirely to developed countries—concluded that there were few empirical regularities in this area that held up through time and space. The one exception that might presume to the status of a law was age-selectivity of migrants. All subsequent work has confirmed this judgment. Whether in developed or developing countries, migrants to urban places are concentrated in the young adult years. The distribution may be more or less peaked according to a given country or city but the general configuration remains. For example, the U.N. Mysore Study (31, p. 177) shows that for all in-migrant male household heads to Bangalore City, 49.9 percent were within the age range 15 to 29 at the time of their migration to that city. The comparable figure for Mysore City is 44.4 percent. Camisa (32, pp. 408-449) applied the survival ratio technique to the 1950 and 1960 censuses of a number of Latin American countries. Six large cities showed the following percentages of net migration represented by ages up to 30 (in 1960):
Greater Buenos Aires 53%
Metropolitan Caracas 67
Greater Santiago 69
Panama City 73
Mexico City (Distrito Federal) 81
Only in Buenos Aires are "older" migrants important as a percentage of theban employment force these migrants to Indian cities to maintain close ties with their folks at their native places." Gutkind (29) reviewing the evidence for Africa, takes much the same position. In addition to fluctuations in demand for labor, the frequent inadequacies of urban facilities (housing, transportation, etc.) make life difficult. The conclusion to be drawn from this is that if economic opportunity and living conditions in the cities were to be improved, return migration rates would he lower.an, et al. (1 3, Ch. 1). Substantively, Ghana differs from Mexico in the predominantly male rural-urban migration, in the multilingual character of the society, in the existence of polygamy, and